Diabetes mellitus (or diabetes) is one of the most prevalent diseases in the world today. Diabetes patients have been divided into two classes, namely type I or insulin-dependent diabetes mellitus and type II or non-insulin dependent diabetes mellitus (NIDDM). Non-insulin-dependent diabetes mellitus (NIDDM) accounts for approximately 90% of all diabetics and is estimated to affect 12-14 million adults in the U.S. alone (6.6% of the population). NIDDM is characterized by both fasting hyperglycemia and exaggerated postprandial increases in plasma glucose levels. NIDDM is associated with a variety of long-term complications, including microvascular diseases such as retinopathy, nephropathy and neuropathy, and macrovascular diseases such as coronary heart disease. Numerous studies in animal models demonstrate a causal relationship between long term complications and hyperglycemia. Recent results from the Diabetes Control and Complications Trial (DCCT) and the Stockholm Prospective Study demonstrate this relationship for the first time in man by showing that insulin-dependent diabetics with tighter glycemic control are at substantially lower risk for development and progression of these complications. Tighter control is also expected to benefit NIDDM patients.
Current therapies used to treat NIDDM patients entail both controlling lifestyle risk factors and pharmaceutical intervention. First-line therapy for NIDDM is typically a tightly-controlled regimen of diet and exercise since an overwhelming number of NIDDM patients are overweight or obese (.apprxeq.67%) and since weight loss can improve insulin secretion, insulin sensitivity and lead to normoglycemia. Normalization of blood glucose occurs in less than 30% of these patients due to poor compliance and poor response. Patients with hyperglycemia not controlled by diet alone are subsequently treated with oral hypoglycemics or insulin. Until recently, the sulfonylureas were the only class of oral hypoglycemic agents available for NIDDM. Treatment with sulfonylureas leads to effective blood glucose lowering in only 70% of patients and only 40% after 10 years of therapy. Patients that fail to respond to diet and sulfonylureas are subsequently treated with daily insulin injections to gain adequate glycemic control.
Although the sulfonylureas represent a major therapy for NIDDM patients, four factors limit their overall success. First, as mentioned above, a large segment of the NIDDM population do not respond adequately to sulfonylurea therapy (i.e. primary failures) or become resistant (i.e. secondary failures). This is particularly true in NIDDM patients with advanced NIDDM since these patients have severely impaired insulin secretion. Second, sulfonylurea therapy is associated with an increased risk of severe hypoglycemic episodes. Third, chronic hyperinsulinemia has been associated with increased cardiovascular disease although this relationship is considered controversial and unproven. Last, sulfonylureas are associated with weight gain, which leads to worsening of peripheral insulin sensitivity and thereby can accelerate the progression of the disease.
Recent results from the U.K. Diabetes prospective study also showed that patients undergoing maximal therapy of a sulfonylurea, metformin, or a combination of the two, were unable to maintain normal fasting glycemia over the six year period of the study. U.K. Prospective Diabetes Study 16. Diabetes, 1995, 44, 1249-158. These results further illustrate the great need for alternative therapies. Three therapeutic strategies that could provide additional health benefits to NIDDM patients beyond the currently available therapies, include drugs that would: (i) prevent the onset of NIDDM; (ii) prevent diabetic complications by blocking detrimental events precipitated by chronic hyperglycemia; or (iii) normalize glucose levels or at least decrease glucose levels below the threshold reported for microvascular and macrovascular diseases.
Hyperglycemia in NIDDM is associated with two biochemical abnormalities, namely insulin resistance and impaired insulin secretion. The relative roles of these metabolic abnormalities in the pathogenesis of NIDDM has been the subject of numerous studies over the past several decades. Studies of offspring and siblings of NIDDM patients, mono- and dizygotic twins, and ethnic populations with high incidence of NIDDM (e.g. Pima Indians) strongly support the inheritable nature of the disease.
Despite the presence of insulin resistance and impaired insulin secretion, fasting blood glucose (FBG) levels remain normal in pre-diabetic patients due to a state of compensatory hyperinsulinemia. Eventually, however, insulin secretion is inadequate and fasting hyperglycemia ensues. With time insulin levels decline. Progression of the disease is characterized by increasing FBG levels and declining insulin levels.
Numerous clinical studies have attempted to define the primary defect that accounts for the progressive increase in FBG. Results from these studies indicate that excessive hepatic glucose output (HGO) is the primary reason for the elevation in FBG with a significant correlation found for HGO and FBG once FBG exceeds 140 mg/dL. Kolterman, et al., J. Clin. Invest. 1981, 68, 957; DeFronzo, Diabetes 1988, 37, 667.
HGO comprises glucose derived from breakdown of hepatic glycogen (glycogenolysis) and glucose synthesized from 3-carbon precursors (gluconeogenesis). A number of radioisotope studies and several studies using .sup.13 C-NMR spectroscopy have shown that gluconeogenesis contributes between 50-100% of the glucose produced by the liver in the postabsorptive state and that gluconeogenesis flux is excessive (2- to 3-fold) in NIDDM patients. Magnusson, et al. J. Clin. Invest. 1992, 90, 1323-1327; Rothman, et al., Science 1991, 254, 573-76; Consoli, et al. Diabetes 1989, 38, 550-557.
Gluconeogenesis from pyruvate is a highly regulated biosynthetic pathway requiring eleven enzymes (FIG. 1). Seven enzymes catalyze reversible reactions and are common to both gluconeogenesis and glycolysis. Four enzymes catalyze reactions unique to gluconeogenesis, namely pyruvate carboxylase, phosphoenolpyruvate carboxykinase, fructose-1,6-bisphosphatase and glucose-6-phosphatase. Overall flux through the pathway is controlled by the specific activities of these enzymes, the enzymes that catalyzed the corresponding steps in the glycolytic direction, and by substrate availability. Dietary factors (glucose, fat) and hormones (insulin, glucagon, glucocorticoids, epinephrine) coordinatively regulate enzyme activities in the gluconeogenesis and glycolysis pathways through gene expression and post-translational mechanisms.
Of the four enzymes specific to gluconeogenesis, fructose-1,6-bisphosphatase (hereinafter "FBPase") is the most suitable target for a gluconeogenesis inhibitor based on efficacy and safety considerations. Studies indicate that nature uses the FBPase/PFK cycle as a major control point (metabolic switch) responsible for determining whether metabolic flux proceeds in the direction of glycolysis or gluconeogenesis. Claus, et al., Mechanisms of Insulin Action, Belfrage, P. editor, pp.305-321, Elsevier Science 1992; Regen, et al. J. Theor. Biol., 1984, 111, 635-658; Pilkis, et al. Annu. Rev. Biochem, 1988, 57, 755-783. FBPase is inhibited by fructose-2,6-bisphosphate in the cell. Fructose-2,6-bisphosphate binds to the substrate site of the enzyme. AMP binds to an allosteric site on the enzyme.
Synthetic inhibitors of FBPase have also been reported. McNiel reported that fructose-2,6-bisphosphate analogs inhibit FBPase by binding to the substrate site, J. Med. Chem., 1984, 106, 7851; U.S. Pat. No. 4,968,790 (1984). These compounds, however, were relatively weak and did not inhibit glucose production in hepatocytes presumably due to poor cell penetration.
Gruber reported that some nucleosides can lower blood glucose in the whole animal through inhibition of FBPase. These compounds exert their activity by first undergoing phosphorylation to the corresponding monophosphate. EP 0 427 799 B1.
Gruber et al. U.S. Pat. No. 5,658,889 described the use of inhibitors of the AMP site of FBPase to treat diabetes.
European Patent Application EP 0 449 196 A3 discloses certain ethyl phosphonates of indole compounds where X is C2-C3 alkylaminocarbonyl. The publication describes inhibitory activity of bone resorption. There is no suggestion that such compounds could be FBPase inhibitors or used to lower blood glucose.
European Patent Application EP 0 620 227 A1 published on Oct. 19, 1994 discloses certain heterocycles including indoles having a phosphonic acid where the X linker is alkylamino and alkylaminoalkyl. These compounds are said to inhibit bone resorption. There is no suggestion that the disclosed compounds were FBPase inhibitors or that they have blood glucose lowering activity.
Japanese Patent Application JP 5-97883 discloses certain indole phosphonates and phosphonate lower alkyl esters where X is a 1,4 linked alkylaryl group. The reference discloses their use for treating hyperlipemia and diabetes mellitus activity. There is no suggestion that such compounds would be FBPase inhibitors. Furthermore, such compounds would not be suitable FBPase inhibitors because a 1,4 linked 6-membered arylalkyl group would have too great a distance between the phosphorus and the indole ring for activity as an FBPase inhibitor at the AMP site.